back
5 comments
All else being equal, demand for a product increases when the prices of its complements decrease.

For the first time, developers can use GitHub for their private projects with up to three collaborators per repository for free.

The free repos are kind of a loss leader to bring in users onto the platform to keep Github as the de facto provider.

This is great news. I still prefer then issue management abilities of Gitlab nested projects, but choice and competition are good for the end user.

Very interesting article. It could be argued that the webapps/internet/browsers are now doing the same to the OS. For casual uses any OS is the same.
Yes, Microsoft realized that long ago. Gates and Balmer prolonged the inevitable through monopoly power, but Nadella has accepted the future and is "turning the aircraft carrier."

It's amazing that Microsoft came back to life from being a zombie company. In hindsight, it was such a large zombie that it had sufficient time to change business models. Somehow Sears wasn't. There's a business school case study and tenure buried in there.

If by 'zombie company' you mean raking in massive amounts of money and growing spectacularly every year? [1]

MSFT has changed in some positive ways, but since Satya took over, growth has been actually flat-ish to negative overall until 2018 (a good year). Paradoxically, MSFT stock was flat during the Ballmer years when there was rock solid consistent growth.

The 'new MSFT' things is quite overrated in terms of what it really means for the company overall. We HN readers tend to considerably overvalue actions that might seem positive, such as things like open source contributions. MS VS Code is great, but it's not hugely material next to core business. Most of the world really does not care, and they are buying gadzooks worth of MSFT products and services. For better or worse.

[1] https://www.statista.com/statistics/267805/microsofts-global...

When I last worked there and ballmer was running things, he thought he was doing great cause he got more money each year from stupid enterprise agreements with stupid companies. But the innovation was really slow and it was all about increasing the value of what they had, windows. The world was moving on without them.

Satya nadella made a surprising change for Microsoft. If a company is going to last long term they have to do more than look at their own navels. You can't just focus on squeezing more money out of your existing lemons.

Happening now at Apple?

Look, as an ex-MSFT person I'm only glad to see someone shit on Ballmer. I didn't like the guy myself. But the pillars of today's Microsoft success were all created by him. Azure? Ballmer. Subscription services (and Office 365 in particular)? Ballmer. XBOX? Ballmer. Satya gets _way_ too much credit for what Microsoft is today. I feel that's mostly because he looks better on stage, and that's a very low bar. A raging gorilla looks better on stage than Ballmer.
Can you show me that Ballmer created XBOX?
The nerve of the CEO growing revenue and profit! When will they ever learn from the unicorns - success is all about revenue growth, not being profitable and going public.
There were several good responses to my message like yours.

A ceo of a tech company has to both grow revenue and cultivate new technologies. A major company working in almost every area like microsoft has to do even more, so they don't get usurped like endless other former leading companies. Microsoft was challenged in a lot of ways. Ballmer did very well on the bringing in more money area.

1. mobile phones (basically completely blew it here - they had all the pieces but couldn't put them together because of both poor execution and more nimble competitors). Ballmer did terrible here. Even buying nokia was a disaster.

2. web - they managed to keep it microsoft centric for a long time where it only rendered correctly on ie3, but it eventually escaped them. they do sell some web server software, but the days where it had to render correctly on windows, both preventing other platforms from being used and being used for hosting web apps are gone. He held the world at bay for a long time, so I'll give him 25% success here.

3. cloud - azure started under him, so I'd give him 50% success. Who would have thought amazon would kick Microsoft's butt so thoroughly. I saw so many good people go to work there, I never thought amazon would succeed with such a different philosophy, people leaving there so frequently after 2 years.

4. Money, growth - 4/4. So much money. Yet at the same time, people started looking for other companies, and the company felt more and more like it was missing out by forcing everything to reinforce windows and office.

Satya has at least changed the tenor of the company, attracting developers back. I do think Windows seems to have turned into a terrible mess, with unless beta releases with lots of bugs. They fired all the testers because they weren't doing anything? Satya was also leading the azure division, so I think he does get some credit for that.

If Microsoft had “won” in Mobile by licensing Windows Mobile, where would it have gotten them? Google “won over” OEMs and got a whopping $21 billion in profit over eight years for their trouble.

https://www.theverge.com/2016/1/21/10810834/android-generate...

There was speculation that Microsoft made more from Android for years from patents than Google made.

Besides, Microsoft is still making pretty good money on mobile by targeting developers with Azure and tools like Xamarin

That's great but you haven't provided any specific examples.

Ballmer did not 'squeeze lemons' - he multiplied the company by like 4x. The 'cloud' FYI was a big new thing. All MS products have evolved quite a lot and are not the 'same products'.

What new products / initiatives has Satya introduced that are lining up to be the future of MS? That will replace the core products sometime?

I honestly thought MS would die. Of course that was my own bias because I had/have not used Windows/Office in 15+ years. But after it became apparent that web/Mobile was the future (and it didn't matter what OS you were on), and MS showed itself to be terrible at Web/Mobile over and over again (IE/Edge, Windows Phone, etc.), and there were good-enough MS Office replacements (in my view); I just didn't see how they would survive. The Googles/Amazons of the world were clearly the future. But I didn't count on MS coming out with Azure. I don't know anyone personally that uses Azure, but I hear it is in second or third place after AWS, which I think is impressive for what I thought was a zombie company.
I'm not sure what you're talking about. The link you sent shows their revenue increasing dramatically since he took over (2014), and their stock shows the same.
Revenue peaked in 2015 when he took over, and only got past that in 2018. So net 'going down' until 2018. Yes, the stock has risen quite dramatically during his time though. Even has revenues went down - wall street dramatically lifted his valuation.

Contrast that with Ballmer, who consistently and reliably grew the company with solid numbers for over a decade, and was rewarded with a stock lift of 0.

The stock lift of MSFT over the last few years is the result of Ballmer. A) his tenure and B) the fact the stock was obviously suppressed a little bit during that time, only recently have the multiples been allowed to shift, and when they did ... the valuation jumped up.

Satya seems to be doing well, but for now it's mostly politics and perception. We'll have to wait and see if his material contributions lead to material change in terms of product and revenue at MSFT.

Ballmer left a company that was a mess. He was boosting profits through repeated layoffs and price squeezing. It had just bought my former employer (Nokia's phone business) without a clear strategy. Windows Phone was already failing in the market due to a series of strategic blunders, arrogance, and poor product positioning and platform strategy. Windows 7 and 8 had largely flopped in the market with lackluster adoption for the same reason. The cloud business was heavily windows focused and not an easy sell either: same reason. Meanwhile developers were jumping ship in favor of Apple laptops and Linux servers, android/ios phones, and standards based web applications on top of Chrome and Firefox.

I grew up in the nineties; things were different then. Visual Studio was the go to product for many developers and developing on windows was a given. Macs didn't become useful as developer machines until a decade later and Linux was a something you ran in your closet or on a old hand me down PC (been there done that). That changed under Ballmer's rule and he failed to adapt. He alienated developers and failed to move windows forward; all while bashing Linux and the Open source community in futile attempts to stem the bleeding of mindshare, revenue, and share holder value.

Fast forward five years or so since Nadella took over and the stock has responded positively to post Ballmer changes by Nadella to bring MS into this century after what was a lost decade.

MS now does open source and established itself as a credible player in the OSS community, it killed off windows phone, opened up the lucrative office ecosystem to non windows platforms by e.g. treating Android and IOS as first class citizens and supporting things like Chrome OS. By embracing linux and open source, and by turning Azure into a very competitive product that mostly runs Linux. He just took Edge out the back and shot it as well.

The net result of all this is that he re-established MS as a company that appeals to developers because despite his famous Mantra, developers were running away under Ballmer. VS Code, having SQL Server supported on linux, supporting Linux and contributing to it, embracing the wider C# ecosystem outside of the Windows silo, buying Github, etc. are all clever moves that rebuilt lost trust. All these things were unthinkable under Ballmer. Even Windows 10 is starting to look pretty nice to people like me with it's linux subsystem, Docker support, and nice hardware options.

As a consequence they have a lot of healthy growth in their cloud business because developers take them serious again. Shareholders are liking this and their prospects for more growth through a portfolio that is no longer just wintel laptop based revenue.

"Ballmer left a company that was a mess. He was boosting profits through repeated layoffs and price squeezing."

I see no evidence for your statement.

Ballmer 4x-ed MSFT revenue during his tenure.

He made far more money for MSFT than Gates ever did, and his growth track record is better than Satya so far.

Ballmer pushed XBox, Azure, Office/Cloud, re-orged the sales teams. Increased revenues for all the other major product lines.

Bing and Mobile were a fail, but they did other things well, moreover, they're up against the top 2 companies in the world in Search and Mobile - no easy feat.

All of your points are descriptive, but they don't really represent huge materiality in terms of business outcomes.

MS was never losing developers.

It's 'OSS contributions' really don't make that much of a difference.

Revenue under Satya has been up and down, with no real progress for his first several years until 2018 (2015 being the watermark).

MSFT today was made by Ballmer, most of the things you indicated are good, but not hugely fundamental.

It’s not an apples to apples comparison.

Growth and/or stock market performance are often only indirectly coupled or lag by years better strategy, innovation, management, etc.

Unfortunately sometimes good decisions never cause growth, there’s no simple formula.

Sears seemed to be intentionally run into the ground so that Eddie Lampert could loot the assets:

https://prospect.org/article/how-sears-was-gutted-its-own-ce...

Whatever bad things you can say about Ballmer, he wasn't out to gut Microsoft. He simply made decisions that were short-sighted and insular. Well, and the review system (which caused a bunch of good people to leave for Google in Kirkland, and some other places).

Once, I was doing an embedded system for a consumer product in the Xbox group. We started out trying to using Windows CE, and after three months of pushing that horrible mess up a hill [holy Toledo, what a steaming pile...] we ditched it in favor of something really simple that could actually do the job, instead of the politically correct thing that we'd been "encouraged" to do. (Honestly, words fail me at how badly designed and buggy the the whole WinCE code base was).

At the time, Ballmer was all about everything having to be Windows. "Windows Foundation" for TV, "Windows Auto", Windows This-and-That. I joked that we wouldn't be allowed to ship our firmware until we named it something like "Windows Just An Event Loop And Some Interrupt Handlers".

I don't know, the windows UI since Windows 10 is a mess; the Windows client platform doesn't seem to have its best time.

Seems strange to me that they are neglecting the client platform (and developer tools to! used to be very important under Balmer), the lock in effect used to be driving demand towards NT server.

to put it frankly: I don't quite understand the grand strategy of MS.

They named it wince, ‘nuff said.
Wait, so are you telling us he did in-fact went out to gut Microsoft via his arrogance?
Ballmer had some bad ideas, but he wanted Microsoft to succeed. He didn't see the attrition as a problem.

The Sears folks are pirates, they don't give a shit about Sears and just want to extract as much money as they can.

Big difference.

I've been browsing HN for maybe 6 years and haven't had the pleasure of seeing this article come up, thanks so much for sharing! Few articles have given me such a feeling of 'enlightenment'.
The problem is - hosting repos, especially private ones, is not a 'complementary service' for Github. It's the core of their offer.

Edit: example - Google paying for Chrome, Android - this is 'complementary'. Google giving away free personal ads on search ... 'core'. MSFT giving away free dev tools, Bing search - complementary. Giving away 'Windows Home' for free ... 'core'.

Github's core offering is not small private repos with max 3 contributors.
But it isn't a complement either. This is a classic loss leader.
Having my personal projects on the same platform as my professional ones is complementary.
So is having cheap milk at the same grocery store you get bread at.
Interesting article.

Any literature/thoughts on when to commoditize complements rather than trying to vertically integrate? Private repos might be a complement to github's other business, but it doesn't naturally follow that they should commoditize it in all situations

You’d get a lot of value from reading an old but still relevant book from 1998: Information Rules: A Strategic Guide to the Network Economy by Hal Varian & Carl Shapiro that discusses several of these strategies.

“Commoditize your complements” is a phrase popularized by Joel Spolsky but the book uses the broader of term: “maximize your value”. Commoditization is one of a few ways to maximize your value.